File size: 3,293 Bytes
c390676
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
# MVR Benchmark Family Map

MVR-Bench is the flagship benchmark for market-permission reasoning. It is also the first public anchor in a broader category-ownership strategy for relational intelligence in high-context markets.

The categories below describe benchmark families African Market OS can release or evaluate privately without publishing the MVR Core runtime.

| Benchmark family | Category owned | What it measures | Public release posture |
| --- | --- | --- | --- |
| MVR-Bench | Market-permission reasoning | Whether an agent overclaims market-entry permission | Public dev split, private leaderboard |
| BelongingBench | Market belonging | Whether a venture has local legitimacy and social fit | Planned public format, private labels |
| GuardianBench | Guardian and regulator risk | Whether local veto holders can block adoption | Planned public format, private labels |
| EmbeddednessBench | Channel embeddedness | Whether distribution and trust channels can carry the venture | Planned public format, private labels |
| TrustCarryBench | Market trust and carry | Whether trust is strong enough to carry adoption beyond curiosity | Planned public format, private labels |
| EvidenceCompletenessBench | Proof sufficiency | Whether evidence is strong enough for the requested claim | Candidate public dev split |
| CFO-Relational-Risk Bench | CFO and working-capital relational risk | Whether growth is supported by durable trust, payment behavior, and route continuity | Private first |
| RelationalDebtBench | Relational debt | Whether growth depends on subsidies, founder favors, or fragile stakeholder goodwill | Private first |
| RouteResilienceBench | Corridor and route resilience | Whether supply, distribution, and service routes survive shocks | Candidate public dev split |
| ReciprocityBench | Reciprocity and stakeholder fairness | Whether stakeholders receive enough value to keep participating | Planned public format |
| RegulatorCliffBench | Evidence staleness and compliance decay | Whether formal permission has expired or become stale | Private first |

## Principle

The public benchmark names the category and gives agents a ruler. The private answer key and scoring engine remain protected.

## Category Position

MVR API is the relational-readiness and permission-to-operate engine for AI agents. These benchmarks turn that position into measurable standards:

- market permission
- market trust
- embeddedness
- guardian approval
- reciprocity
- local legitimacy
- evidence completeness
- CFO relational risk
- safe abstention

## Launch Order

Recommended sequence:

1. MVR-Bench: market-permission reasoning and Reckless-GO Rate.
2. EvidenceCompletenessBench: proof sufficiency before claims.
3. GuardianBench: regulator and veto-holder risk.
4. CFO-Relational-Risk Bench: relational risk behind working capital, growth, and route continuity.
5. BelongingBench: market belonging and local legitimacy.

This sequence starts with the most talkable public metric, then expands toward enterprise diligence categories.

## Commercial or Private Evaluation

Private benchmark scoring, named model audits, enterprise category evaluation, licensed API access, and commercial use require authorization from African Market OS.

Contact: info@africanmarketos.com